Noninvasive Vascular Assessment in Suspected Acute Basilar Artery Occlusion
To the Editor:
Brandt and colleagues1 concluded in their recent article in Stroke that CT angiography (CTA) was superior to Doppler sonography (DS) in the assessment of occlusion or patency of the basilar artery (BA) and that the usefulness of combined extracranial and transcranial DS was confirmed in proximal BA occlusion. I agree with the first conclusion but disagree with the second one and would like to raise the following comments.
First, digital subtraction angiography (DSA), the gold standard for diagnosis or exclusion of BA occlusion, was performed in only 6 of the 19 patients.1 Although I understand the reasons for not performing DSA in all patients, the absence of the gold standard does not allow a direct comparison between DSA and CTA, nor does it permit full validation of the accuracy of CTA in the assessment of patients with suspected BA occlusion.
Second, I am concerned about the accuracy of the “final diagnosis” of some patients. The method of deriving the final diagnosis was not explained in the Subjects and Methods section.1 Transient BA occlusion was implied in 3 patients when infarcts in the posterior circulation were seen on the follow-up CT of the brain. Nevertheless, CT is inferior to MRI in revealing small infarcts, especially in the posterior fossa. I wonder whether MRI was performed in the other patients (1, 3, 4, 12, 16, and 18) to exclude small posterior circulation infarcts which, in turn, would indicate transient BA occlusion. On the other hand, transient BA occlusion should have been diagnosed in patient 7, who had a lower brain stem infarct. Patient 17 did not have DS before CTA; heparin was listed as the treatment, and DSA was listed as not performed in the Table,1 but DSA and local thrombolysis were done, according to the text in the Subjects and Methods and the Results sections.
Third, CTA was performed in 20 patients with suspected BA occlusion, and 1 patient was excluded because DSA was not done. If one adopts the authors’ clinicoradiological diagnostic criteria for BA occlusion, 10 patients with complete BA occlusion and 2 with incomplete BA occlusion were detected by CTA; 1 patient with BA occlusion was missed by CTA because of “dense” BA calcification, and patient 7 was also missed (as explained above). The sensitivity, specificity, positive predictive value, and negative predictive value became 12/14 (85.7%), 6/6 (100.0%), 12/12 (100.0%), and 6/8 (75.0%), respectively, which indicates the usefulness of CTA in patients with suspected BA occlusion. In addition to the limitations of CTA, which were described by the authors,1 the extra dose of nonionic contrast medium may limit the volume of contrast permitted for DSA and/or local intra-arterial thrombolysis.
Fourth, I am interested to know the CT number over the hyperdense BA of patient 19 (Figure 1),1 since the differential consideration is a “hyperdense” BA from thrombosis.2 Finally, the potential role of DS in BA occlusion appears to be very limited, because findings of DS were not helpful in almost half (9/19 or 47.4%) of the target patients with suspected BA occlusion.1 More importantly, time is a critical factor in acute therapy for ischemic stroke.3 Precious time will be lost in attempting to complete the extracranial and transcranial DS when 5% to 19% of patients have poor temporal bone windows,4 findings of DS are not helpful nearly half of the time,1 and BA occlusion cannot be excluded by “normal” DS findings.1
- Copyright © 1999 by American Heart Association
Brandt T, Knauth M, Wildermuth S, Winter R, von Kummer R, Sartor K, Hacke W. CT angiography and Doppler sonography for emergency assessment in acute basilar artery ischemia. Stroke.. 1999;30:606–612.
Leys D, Pruvo JP, Godefroy O, Rondepierre P, Leclerc X. Prevalence and significance of hyperdense middle cerebral artery in acute stroke. Stroke.. 1992;23:317–324.
Saver JL, Feldmann E. Basic transcranial Doppler examination: technique and anatomy. In: Babikian VL, Wechsler LR, eds. Transcranial Doppler Ultrasonography. St Louis, Mo: CV Mosby Co; 1993:11–28.
We appreciate the detailed comments of Dr Cheung, which allow us to further underscore some aspects of our study.
As there exists no prospective study of CTA controlled by DSA in patients suspected of BA occlusion, we published a series of such patients who all had undergone CTA as well as DS to analyze the CTA potential in emergencies.R1 For ethical reasons DSA was performed only on patients considered suitable for thrombolysis. We concluded cautiously that—although we found CTA a reliable non-invasive technique in showing BA occlusion—“the small number of patients studied precludes definite conclusion on the accuracy of CTA in BA occlusion” and that “further evaluation of CTA especially in high-grade BA stenosis is necessary.” Further, we avoided to determine sensitivity and specificity which can only be done in trials with DSA in all patients.
Since the end of our study in February 1997, we and others continued to find CTA reliable in the diagnosis of both BA occlusions and arterial occlusions in the arterior circulation.R2 R3 In 16 additional patients who underwent DSA and intra-arterial thrombolysis for BA occlusion following CTA, diagnostic accuracy was similar. Moreover, Dr Cheung, while correctly calling DSA the “gold standard,” did not point to any diagnostic failure of CTA in his own experience.
Regarding the use of Doppler sonography (DS) in the emergency assessment of suspected BA occlusion, we are not aware of any larger series. We concluded that “the decision not to perform DSA is not to be based reliably on negative DS findings alone if BA occlusion is clinically suspected.”R1 However, this does not preclude any use of DS in BA occlusion. Indeed, we found positive results by DS in 3 of 4 proximal BA occlusions. With such positive signs for BA occlusion as a high-resistance pattern, sudden loss of signal, or retrograde flow in the distal BA, diagnosis of BA occlusion can be made. BA occlusion can be excluded with a high degree of certainty if there are normal DS findings, including absence of indirect signs for BA occlusion such as pathological flow patterns in the vertebral or posterior cerebral arteries and a sufficient insonation depth of the BA of at least 110 to 120 mm by transcranial Doppler sonography.R4 We encountered false-negative results or uncertain DS findings of BA occlusion only in patients with an insonation depth below these values or due to technical problems (eg, intubation, adipose necks); in these cases, further imaging studies became necessary. Both patients in our study with false-negative DS results and distal BA occlusion had a maximal insonation depth of only 90 and 100 mm, respectively. Furthermore, a particular strength of DS is the assessment of hemodynamics, which can also be used in monitoring indirect signs of BA occlusion. Thus, DS may complement the information gained by CTA in the decision for thrombolytic therapy. In fact, only recently we treated by intra-arterial thrombolysis a patient in good clinical condition with a short distal BA occlusion diagnosed by CTA, because DS revealed a severe flow reduction in the BA and both posterior cerebral arteries. Moreover, the vascular region of interest must be known before CTA is performed, as the scanning range is relatively narrow. DS further assists—along with the clinical findings—in identification of this region (ie, extracranial versus intracranial, anterior versus posterior circulation). The diagnostic value of DS and the time spent for the examination strongly depend on the experience of the examiner, though; experienced examiners need no more than 10 minutes if the clinically relevant arteries are focused on. By contrast, CTA is less dependent on the experience of the examiner.
Definition and accuracy of “final diagnosis”: All patients were emergency patients initially suspected clinically of having acute BA occlusion as defined in the Methods section.R1 Not all patients were eventually admitted to our maximal-care hospital. Therefore, or due to poor clinical condition, MR imaging could not be performed in some patients. The final diagnosis was defined by clinical course and final imaging results. In most cases brain stem ischemia caused by BA occlusion or associated with posterior circulation infarcts shown by MRI in some of these patients (2, 5, 6, 7, 8, 13, 14, and 15) or by follow-up CT was the final diagnosis. In patient 7, a lower brain stem infarct in the territory of the posterior cerebellar artery was found on MRI, which made a transient BA occlusion highly unlikely. There were, however, 3 patients without BA occlusion and brain stem infarct even on follow-up imaging, in whom could be identified other etiologies, such as intoxication (patient 18, proven by laboratory test), postictal coma, and a large hemispheric infarct detected on follow-up CT.
In patient 17 with BA occlusion, no DSA was performed because of thalamic and occipital lobe infarcts and age (77 years). The CT number (in Hounsfield units) over the hyperdense BA of patient 19 (Figure 1)R1 was 280, which corresponds to calcification. Finally, as stated in our previous article of CTA in the anterior circulation, the use of 130 mL of nonionic contrast agent for CTA does not preclude subsequent DSA.R2
Brandt T, Knauth M, Wildermuth S, Winter R, Sartor K, von Kummer R, Hacke W. CT angiography and Doppler sonography for the emergency assessment in acute basilar artery ischemia. Stroke.. 1999;30:606–612.
Wildermuth S, Knauth M, Brandt T, Winter R, Sartor K, Hacke W. Role of computerized tomography angiography in patient selection for thrombolytic therapy in acute hemispheric stroke. Stroke.. 1998;29:935–938.
Knauth M, von Kummer R, Jansen O, Hähnel S, Dörfler A, Sartor K. Potential of CT angiography in acute ischemic stroke. AJNR Am J Neuroradiol.. 1997;18:1001–1010.